Alcohol Advertising Risk-taking Studies
Study 1 (4a)
Descriptives and Chart
| Prime Type | Risk Response | Proportion | SD | N |
|---|---|---|---|---|
| Beer | High | 0.564 | 0.181 | 112 |
| Beer | Low | 0.390 | 0.168 | 112 |
| Neutral | High | 0.482 | 0.180 | 112 |
| Neutral | Low | 0.459 | 0.164 | 112 |
| Water | High | 0.502 | 0.183 | 112 |
| Water | Low | 0.454 | 0.168 | 112 |
Charts including condition
DV: Proportion of high risk responses - proportion of low risk responses
DV: Risk variable: number of high risk responses/(number of high risk + low risk responses)
Models
Simple Effects
Tests if the difference in propotions of high-risk responses is different from the proportions of low-risk for each prime category.
Beer
##
## Call:
## lm(formula = beer.dif ~ 1, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -24.1696 -4.4196 0.8304 5.8304 17.8304
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.1696 0.7823 5.33 5.21e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 8.279 on 111 degrees of freedomWater
##
## Call:
## lm(formula = water.dif ~ 1, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -20.1518 -4.1518 0.8482 5.8482 17.8482
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.1518 0.7861 1.465 0.146
##
## Residual standard error: 8.319 on 111 degrees of freedomNeutral
##
## Call:
## lm(formula = neut.dif ~ 1, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -19.5357 -4.7857 0.4643 5.4643 23.4643
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.5357 0.7654 0.7 0.485
##
## Residual standard error: 8.101 on 111 degrees of freedomMixed Effects Models
Beer vs Water
This model tests the effects of a Beer vs a Water prime on risk response.
Fixed effects:
      Beer/Water contrast: Beer = +1/2, Water = -1/2, Neutral = 0
      Orthogonal contrast: Beer = +1/3, Water = +1/3, Neutral = -2/3
Random effects:
      Subject: Intercept, slope for each contrast
      Stimuli(pic2): Intercept
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 10174.6 10244.1 -5077.3 10154.6 7652
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3901 -0.9605 0.5990 0.8288 2.5831
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.31142 0.5581
## BvW 0.64767 0.8048 -0.11
## BWvN 0.29586 0.5439 -0.05 0.32
## pic2 (Intercept) 0.01092 0.1045
## Number of obs: 7662, groups: Subject, 112; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.16979 0.06548 2.593 0.00951 **
## BvW 0.30029 0.12171 2.467 0.01361 *
## BWvN 0.20470 0.09689 2.113 0.03462 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW
## BvW -0.054
## BWvN -0.021 0.110This model adds the Alcohol Quantity-Frequency Product (AlcQF) and its interactions to the list of fixed effects and calculates an AlcQF slope for each stimuli picture.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + AlcQF * BvW + AlcQF * BWvN + (1 + BvW +
## BWvN | Subject) + (1 + AlcQF | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 9638.4 9741.7 -4804.2 9608.4 7245
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4242 -0.9552 0.5908 0.8302 2.6244
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 3.188e-01 0.564630
## BvW 6.405e-01 0.800337 -0.10
## BWvN 2.900e-01 0.538481 -0.01 0.43
## pic2 (Intercept) 1.290e-02 0.113569
## AlcQF 1.541e-05 0.003926 -0.30
## Number of obs: 7260, groups: Subject, 106; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.167531 0.068664 2.440 0.0147 *
## BvW 0.282476 0.127581 2.214 0.0268 *
## BWvN 0.224964 0.101843 2.209 0.0272 *
## AlcQF -0.006184 0.005952 -1.039 0.2988
## BvW:AlcQF 0.004975 0.009971 0.499 0.6178
## BWvN:AlcQF 0.008492 0.007618 1.115 0.2649
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW BWvN AlcQF BW:AQF
## BvW -0.045
## BWvN -0.003 0.138
## AlcQF -0.025 -0.001 -0.001
## BvW:AlcQF -0.001 -0.049 -0.001 -0.067
## BWvN:AlcQF -0.002 -0.001 -0.058 -0.009 0.217Beer vs Neutral
This model tests the effects of a Beer vs a Neutral prime on risk response.
Fixed effects:
      Beer/Neutral contrast: Beer = +1/2, Neutral = -1/2, Water = 0
      Orthogonal contrast: Beer = +1/3, Neutral = +1/3, Water = -2/3
Random effects:
      Subject: Intercept, slope for each contrast
      Stimuli(pic2): Intercept
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 10174.6 10244.1 -5077.3 10154.6 7652
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3900 -0.9605 0.5990 0.8288 2.5831
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.31142 0.5580
## BvW 0.64766 0.8048 -0.11
## BWvN 0.29585 0.5439 -0.05 0.32
## pic2 (Intercept) 0.01092 0.1045
## Number of obs: 7662, groups: Subject, 112; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.16978 0.06548 2.593 0.00951 **
## BvN 0.35487 0.11996 2.958 0.00309 **
## BNvW 0.12286 0.09850 1.247 0.21230
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN
## BvN -0.044
## BNvW -0.039 0.128This model adds the Alcohol Quantity-Frequency Product (AlcQF) and its interactions to the list of fixed effects and calculates an AlcQF slope for each stimuli picture.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + AlcQF * BvN + AlcQF * BNvW + (1 + BvN +
## BNvW | Subject) + (1 + AlcQF | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 9638.4 9741.7 -4804.2 9608.4 7245
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4242 -0.9552 0.5908 0.8302 2.6244
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 3.188e-01 0.564629
## BvN 6.359e-01 0.797464 -0.06
## BNvW 2.934e-01 0.541663 -0.10 0.44
## pic2 (Intercept) 1.290e-02 0.113568
## AlcQF 1.541e-05 0.003925 -0.30
## Number of obs: 7260, groups: Subject, 106; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.1675359 0.0686622 2.440 0.01469 *
## BvN 0.3662160 0.1274089 2.874 0.00405 **
## BNvW 0.0993781 0.1019915 0.974 0.32987
## AlcQF -0.0061835 0.0059515 -1.039 0.29881
## BvN:AlcQF 0.0109790 0.0099699 1.101 0.27081
## BNvW:AlcQF -0.0005151 0.0076182 -0.068 0.94609
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN BNvW AlcQF BN:AQF
## BvN -0.025
## BNvW -0.041 0.140
## AlcQF -0.025 -0.002 0.000
## BvN:AlcQF -0.002 -0.048 -0.002 -0.040
## BNvW:AlcQF 0.000 -0.002 -0.059 -0.061 0.218Water vs Neutral
This model tests the effects of a Water vs a Neutral prime on risk response.
Fixed effects:
      Water/Neutral contrast: Water = +1/2, Neutral = -1/2, Beer = 0
      Orthogonal contrast: Water = +1/3, Neutral = +1/3, Beer = -2/3
Random effects:
      Subject: Intercept, slope for each contrast
      Stimuli(pic2): Intercept
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 10174.6 10244.1 -5077.3 10154.6 7652
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3900 -0.9605 0.5990 0.8288 2.5831
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.31142 0.5581
## BvW 0.64766 0.8048 -0.11
## BWvN 0.29585 0.5439 -0.05 0.32
## pic2 (Intercept) 0.01092 0.1045
## Number of obs: 7662, groups: Subject, 112; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.16978 0.06548 2.593 0.00951 **
## WvN 0.05458 0.10858 0.503 0.61523
## WNvB -0.32758 0.10796 -3.034 0.00241 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN
## WvN 0.011
## WNvB 0.055 0.018This model adds the Alcohol Quantity-Frequency Product (AlcQF) and its interactions to the list of fixed effects and calculates an AlcQF slope for each stimuli picture.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + AlcQF * WvN + AlcQF * WNvB + (1 + WvN +
## WNvB | Subject) + (1 + AlcQF | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 9638.4 9741.7 -4804.2 9608.4 7245
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4242 -0.9552 0.5908 0.8302 2.6244
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 3.188e-01 0.564626
## WvN 2.643e-01 0.514054 0.07
## WNvB 5.722e-01 0.756439 0.08 0.01
## pic2 (Intercept) 1.290e-02 0.113568
## AlcQF 1.542e-05 0.003926 -0.30
## Number of obs: 7260, groups: Subject, 106; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.167538 0.068662 2.440 0.01469 *
## WvN 0.083729 0.112455 0.745 0.45654
## WNvB -0.324342 0.114429 -2.834 0.00459 **
## AlcQF -0.006183 0.005951 -1.039 0.29881
## WvN:AlcQF 0.006005 0.008147 0.737 0.46105
## WNvB:AlcQF -0.007977 0.009100 -0.876 0.38075
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN WNvB AlcQF WN:AQF
## WvN 0.023
## WNvB 0.039 0.002
## AlcQF -0.025 -0.001 0.002
## WvN:AlcQF -0.001 -0.065 -0.001 0.032
## WNvB:AlcQF 0.002 -0.001 -0.045 0.058 0.000Models including condition
Mixed effects model
This mixed model tests a contrast between risks decisions made after Beer primes vs Water and Neutral primes combined (WNvB). It includes video group and beverage contrasts and their interaction.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WNvB + WvN + vidCvM * WNvB + vidBvW * WNvB + vidint *
## WNvB + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 10181.2 10292.3 -5074.6 10149.2 7646
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3325 -0.9591 0.5911 0.8323 2.5800
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.3006 0.5482
## BvW 0.6341 0.7963 -0.14
## BWvN 0.2930 0.5413 -0.07 0.32
## pic2 (Intercept) 0.0109 0.1044
## Number of obs: 7662, groups: Subject, 112; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.17015 0.06477 2.627 0.00862 **
## WNvB -0.32999 0.10746 -3.071 0.00213 **
## WvN 0.05440 0.10855 0.501 0.61623
## vidCvM 0.03190 0.05755 0.554 0.57940
## vidBvW 0.03355 0.05730 0.585 0.55828
## vidint 0.09612 0.05732 1.677 0.09359 .
## WNvB:vidCvM -0.03825 0.08660 -0.442 0.65871
## WNvB:vidBvW -0.08057 0.08631 -0.934 0.35056
## WNvB:vidint -0.07271 0.08637 -0.842 0.39990
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WNvB WvN vidCvM vidBvW vidint WNB:CM WNB:BW
## WNvB 0.069
## WvN 0.013 0.018
## vidCvM 0.032 0.002 -0.002
## vidBvW 0.013 -0.002 0.001 -0.018
## vidint -0.015 -0.006 -0.001 0.014 0.036
## WNvB:vidCvM 0.002 0.029 0.001 0.098 -0.007 -0.002
## WNvB:vidBvW -0.002 0.013 0.000 -0.007 0.097 0.003 -0.011
## WNvB:vidint -0.007 -0.009 -0.001 -0.002 0.003 0.097 0.016 0.036Standard analyses (ANOVA Ws)
These models test a Beer prime vs a Water prime contrast with both video factors included in the model. BvW is calculated as: proportion of high risk responses after beer primes - proportion of high risk responses after water primes. BvWrisk is calculated as: beer risk - water risk. Risk variable: high risk response/(hi + lo risk repsonses) for each prime type.
##
## Call:
## lm(formula = BvW ~ vidCvM + vidBvW + vidint, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.97685 -0.29821 -0.01032 0.31724 1.18981
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.12707 0.04184 3.037 0.00299 **
## vidCvM 0.03342 0.04184 0.799 0.42613
## vidBvW 0.03309 0.04184 0.791 0.43071
## vidint 0.02555 0.04184 0.611 0.54266
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4424 on 108 degrees of freedom
## Multiple R-squared: 0.01468, Adjusted R-squared: -0.01269
## F-statistic: 0.5365 on 3 and 108 DF, p-value: 0.6583##
## Call:
## lm(formula = BvWrisk ~ vidCvM + vidBvW + vidint, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.52511 -0.14784 -0.00435 0.17115 0.67170
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.06760 0.02186 3.093 0.00252 **
## vidCvM 0.01722 0.02186 0.788 0.43250
## vidBvW 0.01649 0.02186 0.754 0.45222
## vidint 0.01394 0.02186 0.638 0.52492
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2311 on 108 degrees of freedom
## Multiple R-squared: 0.01431, Adjusted R-squared: -0.01307
## F-statistic: 0.5226 on 3 and 108 DF, p-value: 0.6677This model tests a Beer prime vs a Neutral prime contrast with both video factors included in the model. BvN is calculated as: proportion of high risk responses after beer primes - proportion of high risk responses after neutral primes. BvNrisk is calculated as: beer risk - neutral risk. Risk variable: high risk response/(hi + lo risk repsonses) for each prime type.
##
## Call:
## lm(formula = BvN ~ vidCvM + vidBvW + vidint, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9361 -0.2353 -0.0246 0.2472 1.0756
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.151576 0.040465 3.746 0.00029 ***
## vidCvM 0.001973 0.040465 0.049 0.96120
## vidBvW 0.033829 0.040465 0.836 0.40499
## vidint 0.028671 0.040465 0.709 0.48014
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4278 on 108 degrees of freedom
## Multiple R-squared: 0.01066, Adjusted R-squared: -0.01682
## F-statistic: 0.3878 on 3 and 108 DF, p-value: 0.762##
## Call:
## lm(formula = BvNrisk ~ vidCvM + vidBvW + vidint, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.5067 -0.1361 -0.0120 0.1435 0.5886
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.080456 0.021454 3.750 0.000286 ***
## vidCvM 0.001588 0.021454 0.074 0.941147
## vidBvW 0.017996 0.021454 0.839 0.403414
## vidint 0.017003 0.021454 0.793 0.429787
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2268 on 108 degrees of freedom
## Multiple R-squared: 0.01182, Adjusted R-squared: -0.01563
## F-statistic: 0.4307 on 3 and 108 DF, p-value: 0.7314This model tests a Water prime vs a Neutral prime contrast with both video factors included in the model. NvW is calculated as: proportion of high risk responses after neutral primes - proportion of high risk responses after water primes. NvWrisk is calculated as: neutral risk - water risk. Risk variable: high risk response/(hi + lo risk repsonses) for each prime type.
##
## Call:
## lm(formula = NvW ~ vidCvM + vidBvW + vidint, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.84414 -0.25309 -0.03726 0.19946 1.33025
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0245040 0.0353450 -0.693 0.490
## vidCvM 0.0314484 0.0353450 0.890 0.376
## vidBvW -0.0007385 0.0353450 -0.021 0.983
## vidint -0.0031195 0.0353450 -0.088 0.930
##
## Residual standard error: 0.3737 on 108 degrees of freedom
## Multiple R-squared: 0.007374, Adjusted R-squared: -0.0202
## F-statistic: 0.2674 on 3 and 108 DF, p-value: 0.8487##
## Call:
## lm(formula = NvWrisk ~ vidCvM + vidBvW + vidint, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.44213 -0.12616 -0.01753 0.10579 0.69875
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.012853 0.018363 -0.700 0.485
## vidCvM 0.015634 0.018363 0.851 0.396
## vidBvW -0.001505 0.018363 -0.082 0.935
## vidint -0.003060 0.018363 -0.167 0.868
##
## Residual standard error: 0.1942 on 108 degrees of freedom
## Multiple R-squared: 0.007, Adjusted R-squared: -0.02058
## F-statistic: 0.2538 on 3 and 108 DF, p-value: 0.8585Simples
BvW is calculated as: proportion of high risk responses after beer primes - proportion of high risk responses after water primes. BvWrisk is calculated as: beer risk - water risk. Risk variable: high risk response/(hi + lo risk repsonses) for each prime type.
BvN is calculated as: proportion of high risk responses after beer primes - proportion of high risk responses after neutral primes. BvNrisk is calculated as: beer risk - neutral risk. Risk variable: high risk response/(hi + lo risk repsonses) for each prime type.
CU Beer
##
## Call:
## lm(formula = BvW ~ 1, data = gamble.w[gamble.w$Group == 1, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.84414 -0.28164 0.03086 0.23920 1.03086
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.21914 0.07651 2.864 0.00817 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3976 on 26 degrees of freedom##
## Call:
## lm(formula = BvWrisk ~ 1, data = gamble.w[gamble.w$Group == 1,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.4359 -0.1497 0.0276 0.1293 0.5400
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.11526 0.03981 2.895 0.00758 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2069 on 26 degrees of freedom##
## Call:
## lm(formula = BvN ~ 1, data = gamble.w[gamble.w$Group == 1, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.84105 -0.23688 0.03395 0.15895 0.78395
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.21605 0.07148 3.022 0.00557 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3714 on 26 degrees of freedom##
## Call:
## lm(formula = BvNrisk ~ 1, data = gamble.w[gamble.w$Group == 1,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.44906 -0.12692 0.00417 0.07136 0.40470
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.11704 0.03735 3.134 0.00424 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1941 on 26 degrees of freedomCU Water
##
## Call:
## lm(formula = BvW ~ 1, data = gamble.w[gamble.w$Group == 2, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.97685 -0.37269 -0.06019 0.35648 1.18981
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.10185 0.09743 1.045 0.305
##
## Residual standard error: 0.5063 on 26 degrees of freedom##
## Call:
## lm(formula = BvWrisk ~ 1, data = gamble.w[gamble.w$Group == 2,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.51882 -0.19504 -0.04249 0.17478 0.67170
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.05439 0.05123 1.062 0.298
##
## Residual standard error: 0.2662 on 26 degrees of freedom##
## Call:
## lm(formula = BvN ~ 1, data = gamble.w[gamble.w$Group == 2, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.88272 -0.23688 0.03395 0.26312 1.07562
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.09105 0.09229 0.987 0.333
##
## Residual standard error: 0.4796 on 26 degrees of freedom##
## Call:
## lm(formula = BvNrisk ~ 1, data = gamble.w[gamble.w$Group == 2,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.46602 -0.12330 0.01455 0.13897 0.58857
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.04704 0.04836 0.973 0.34
##
## Residual standard error: 0.2513 on 26 degrees of freedomMU Beer
##
## Call:
## lm(formula = BvW ~ 1, data = gamble.w[gamble.w$Group == 3, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.89286 -0.31994 -0.03869 0.31548 1.02381
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.10119 0.08699 1.163 0.255
##
## Residual standard error: 0.4603 on 27 degrees of freedom##
## Call:
## lm(formula = BvWrisk ~ 1, data = gamble.w[gamble.w$Group == 3,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.4660 -0.1665 -0.0264 0.1577 0.5250
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.05293 0.04445 1.191 0.244
##
## Residual standard error: 0.2352 on 27 degrees of freedom##
## Call:
## lm(formula = BvN ~ 1, data = gamble.w[gamble.w$Group == 3, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.82143 -0.19643 -0.00893 0.28274 0.84524
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.15476 0.07685 2.014 0.0541 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4066 on 27 degrees of freedom##
## Call:
## lm(formula = BvNrisk ~ 1, data = gamble.w[gamble.w$Group == 3,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.42769 -0.08777 -0.00140 0.13984 0.42014
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.07986 0.03929 2.033 0.052 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2079 on 27 degrees of freedomMU Water
##
## Call:
## lm(formula = BvW ~ 1, data = gamble.w[gamble.w$Group == 4, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.91944 -0.24236 -0.00278 0.32014 0.62222
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.08611 0.07302 1.179 0.248
##
## Residual standard error: 0.3999 on 29 degrees of freedom##
## Call:
## lm(formula = BvWrisk ~ 1, data = gamble.w[gamble.w$Group == 4,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.52511 -0.12738 -0.00435 0.19165 0.31267
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.04783 0.03895 1.228 0.229
##
## Residual standard error: 0.2134 on 29 degrees of freedom##
## Call:
## lm(formula = BvN ~ 1, data = gamble.w[gamble.w$Group == 4, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.93611 -0.29028 -0.06111 0.23056 1.06389
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.14444 0.08122 1.779 0.0858 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4448 on 29 degrees of freedom##
## Call:
## lm(formula = BvNrisk ~ 1, data = gamble.w[gamble.w$Group == 4,
## ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.5067 -0.1378 -0.0353 0.1630 0.5363
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.07788 0.04509 1.727 0.0948 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.247 on 29 degrees of freedomCorrelations
NIAAA Use Measures
These include raw and recoded (.re) values. P-values shown. 
Correlation coefficients shown. 
Combined
beer.dif = hi-risk beer responses - low-risk beer responses
water.dif = hi-risk water responses - low-risk water responses
neutral.dif = hi-risk neutral responses - low-risk neutral responses
P-values shown. 
Correlation coefficients shown. 
Study 2 (4b)
Descriptives and Chart
| Prime Type | Risk Response | Proportion | SD | N |
|---|---|---|---|---|
| Beer | High | 0.574 | 0.180 | 66 |
| Beer | Low | 0.361 | 0.170 | 66 |
| Neutral | High | 0.465 | 0.165 | 66 |
| Neutral | Low | 0.487 | 0.142 | 66 |
| Water | High | 0.477 | 0.200 | 66 |
| Water | Low | 0.473 | 0.180 | 66 |
Models
Simple Effects
Tests if the difference in propotions of high-risk responses is different from the proportions of low-risk for each prime category.
Beer
##
## Call:
## lm(formula = beer.dif ~ 1, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -25.106 -4.106 1.394 5.894 16.894
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5.106 1.008 5.066 3.6e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 8.188 on 65 degrees of freedomWater
##
## Call:
## lm(formula = water.dif ~ 1, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -20.0909 -5.8409 0.9091 7.9091 14.9091
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.09091 1.11102 0.082 0.935
##
## Residual standard error: 9.026 on 65 degrees of freedomNeutral
##
## Call:
## lm(formula = neut.dif ~ 1, data = gamble.w)
##
## Residuals:
## Min 1Q Median 3Q Max
## -18.4697 -4.4697 0.5303 5.2803 14.5303
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.5303 0.8919 -0.595 0.554
##
## Residual standard error: 7.246 on 65 degrees of freedomMixed Effects Models
Beer vs Water
This model tests the effects of a Beer vs a Water prime on risk response.
Fixed effects:
      Beer/Water contrast: Beer = +1/2, Water = -1/2, Neutral = 0
      Orthogonal contrast: Beer = +1/3, Water = +1/3, Neutral = -2/3
Random effects:
      Subject: Intercept, slope for each contrast
      Stimuli(pic2): Intercept
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5934.0 5998.1 -2957.0 5914.0 4482
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4593 -0.9455 0.5729 0.8396 2.3742
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.29755 0.5455
## BvW 1.02897 1.0144 -0.22
## BWvN 0.23781 0.4877 0.27 0.32
## pic2 (Intercept) 0.02976 0.1725
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.13198 0.08963 1.472 0.14090
## BvW 0.53168 0.19219 2.766 0.00567 **
## BWvN 0.30344 0.13873 2.187 0.02873 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW
## BvW -0.108
## BWvN 0.096 0.093These models add questionnaire measures and interactions to the list of fixed effects and calculates a questionnaire measure slope for each stimuli picture.
Measures:
AlcQF - Alcohol use quantity-frequency product (drinks/month)
Risk - Risk taking after alcohol consumption
RAPI - Rutgers alcohol problem index
AE - Alcohol expectancy questionnaire (CEOA)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + AlcQF * BvW + AlcQF * BWvN + (1 + BvW +
## BWvN | Subject) + (1 + AlcQF | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5665.1 5760.6 -2817.6 5635.1 4272
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4848 -0.9404 0.5688 0.8360 2.4141
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 3.180e-01 0.56394
## BvW 1.044e+00 1.02159 -0.24
## BWvN 2.477e-01 0.49772 0.25 0.29
## pic2 (Intercept) 3.703e-02 0.19242
## AlcQF 1.452e-05 0.00381 -1.00
## Number of obs: 4287, groups: Subject, 63; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.122016 0.096099 1.270 0.2042
## BvW 0.499600 0.204643 2.441 0.0146 *
## BWvN 0.282200 0.150150 1.880 0.0602 .
## AlcQF -0.002118 0.009785 -0.216 0.8287
## BvW:AlcQF 0.017343 0.019142 0.906 0.3649
## BWvN:AlcQF -0.003782 0.011753 -0.322 0.7476
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW BWvN AlcQF BW:AQF
## BvW -0.111
## BWvN 0.086 0.078
## AlcQF -0.066 0.002 0.002
## BvW:AlcQF 0.002 -0.093 0.001 -0.176
## BWvN:AlcQF 0.003 0.002 -0.160 0.167 0.168## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + Risk * BvW + Risk * BWvN + (1 + BvW +
## BWvN | Subject) + (1 + Risk | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5934.8 6031.0 -2952.4 5904.8 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.7414 -0.9357 0.5778 0.8355 2.3929
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.277207 0.5265
## BvW 1.039112 1.0194 -0.25
## BWvN 0.243954 0.4939 0.29 0.33
## pic2 (Intercept) 0.029690 0.1723
## Risk 0.008081 0.0899 -0.74
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.13253 0.08788 1.508 0.13153
## BvW 0.53544 0.19258 2.780 0.00543 **
## BWvN 0.30416 0.13902 2.188 0.02868 *
## Risk 0.14559 0.07702 1.890 0.05872 .
## BvW:Risk 0.09767 0.16271 0.600 0.54831
## BWvN:Risk -0.01341 0.10625 -0.126 0.89957
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW BWvN Risk BvW:Rs
## BvW -0.118
## BWvN 0.104 0.096
## Risk -0.139 0.004 0.004
## BvW:Risk 0.004 -0.177 0.004 -0.150
## BWvN:Risk 0.004 0.004 -0.290 0.159 0.157## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + RAPI * BvW + RAPI * BWvN + (1 + BvW +
## BWvN | Subject) + (1 + RAPI | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5939.9 6036.0 -2954.9 5909.9 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.5182 -0.9471 0.5709 0.8410 2.3844
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 2.789e-01 0.528156
## BvW 1.031e+00 1.015490 -0.23
## BWvN 2.385e-01 0.488406 0.29 0.32
## pic2 (Intercept) 2.975e-02 0.172478
## RAPI 1.315e-06 0.001147 -1.00
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.1321050 0.0880409 1.500 0.13349
## BvW 0.5321226 0.1922882 2.767 0.00565 **
## BWvN 0.3035594 0.1387694 2.188 0.02871 *
## RAPI 0.0188077 0.0100236 1.876 0.06061 .
## BvW:RAPI 0.0001852 0.0204797 0.009 0.99279
## BWvN:RAPI -0.0034730 0.0124243 -0.280 0.77984
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW BWvN RAPI BW:RAP
## BvW -0.109
## BWvN 0.103 0.093
## RAPI -0.022 0.004 0.001
## BvW:RAPI 0.004 -0.028 0.003 -0.167
## BWvN:RAPI 0.001 0.004 -0.049 0.189 0.192## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvW + BWvN + AE * BvW + AE * BWvN + (1 + BvW + BWvN |
## Subject) + (1 + AE | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5941.1 6037.2 -2955.5 5911.1 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4611 -0.9414 0.5719 0.8440 2.3828
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 2.945e-01 0.542646
## BvW 1.011e+00 1.005368 -0.24
## BWvN 2.270e-01 0.476497 0.25 0.30
## pic2 (Intercept) 2.974e-02 0.172449
## AE 6.344e-06 0.002519 -1.00
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.131670 0.089362 1.473 0.14063
## BvW 0.531137 0.191458 2.774 0.00553 **
## BWvN 0.303186 0.138123 2.195 0.02816 *
## AE 0.006774 0.008962 0.756 0.44975
## BvW:AE 0.018969 0.017835 1.064 0.28751
## BWvN:AE 0.012118 0.010815 1.121 0.26251
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW BWvN AE BvW:AE
## BvW -0.115
## BWvN 0.089 0.085
## AE -0.042 -0.001 0.001
## BvW:AE -0.001 -0.059 -0.001 -0.170
## BWvN:AE 0.001 -0.001 -0.105 0.163 0.180This model includes all the questionnaire measures described above as fixed effects. Only intercepts were calculated for random factors due to convergence limitations.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula:
## response ~ BvW + BWvN + AlcQF + Risk + RAPI + AE + (1 | Subject) +
## (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5734.1 5791.4 -2858.0 5716.1 4278
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.0243 -0.9639 0.6045 0.8820 1.6984
##
## Random effects:
## Groups Name Variance Std.Dev.
## Subject (Intercept) 0.24057 0.4905
## pic2 (Intercept) 0.02882 0.1698
## Number of obs: 4287, groups: Subject, 63; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.113543 0.085183 1.333 0.18256
## BvW 0.446701 0.143346 3.116 0.00183 **
## BWvN 0.277281 0.123736 2.241 0.02503 *
## AlcQF -0.015656 0.009796 -1.598 0.11001
## Risk 0.141890 0.086950 1.632 0.10271
## RAPI 0.018680 0.011280 1.656 0.09772 .
## AE -0.003394 0.009687 -0.350 0.72610
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvW BWvN AlcQF Risk RAPI
## BvW 0.004
## BWvN 0.003 0.005
## AlcQF 0.014 -0.003 -0.001
## Risk -0.018 0.003 0.002 -0.187
## RAPI -0.019 0.004 0.001 -0.340 -0.287
## AE -0.005 -0.002 -0.001 0.036 -0.432 -0.058Beer vs Neutral
This model tests the effects of a Beer vs a Neutral prime on risk response.
Fixed effects:
      Beer/Neutral contrast: Beer = +1/2, Neutral = -1/2, Water = 0
      Orthogonal contrast: Beer = +1/3, Neutral = +1/3, Water = -2/3
Random effects:
      Subject: Intercept, slope for each contrast
      Stimuli(pic2): Intercept
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5934.0 5998.1 -2957.0 5914.0 4482
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4593 -0.9455 0.5729 0.8396 2.3742
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.29755 0.5455
## BvW 1.02896 1.0144 -0.22
## BWvN 0.23782 0.4877 0.27 0.32
## pic2 (Intercept) 0.02976 0.1725
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.13198 0.08963 1.472 0.14090
## BvN 0.56930 0.17594 3.236 0.00121 **
## BNvW 0.24704 0.15406 1.604 0.10882
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN
## BvN 0.017
## BNvW -0.144 0.202These models add questionnaire measures and interactions to the list of fixed effects and calculates a questionnaire measure slope for each stimuli picture.
Measures:
AlcQF - Alcohol use quantity-frequency product (drinks/month)
Risk - Risk taking after alcohol consumption
RAPI - Rutgers alcohol problem index
AE - Alcohol expectancy questionnaire (CEOA)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + AlcQF.c * BvN + AlcQF.c * BNvW + (1 +
## BvN + BNvW | Subject) + (1 + AlcQF.c | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5665.1 5760.6 -2817.6 5635.1 4272
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4848 -0.9404 0.5688 0.8360 2.4141
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 3.180e-01 0.56394
## BvN 6.574e-01 0.81083 0.00
## BNvW 5.374e-01 0.73305 -0.34 0.58
## pic2 (Intercept) 3.702e-02 0.19242
## AlcQF.c 1.452e-05 0.00381 -1.00
## Number of obs: 4287, groups: Subject, 63; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.122005 0.096100 1.270 0.2042
## BvN 0.531998 0.188200 2.827 0.0047 **
## BNvW 0.233607 0.165518 1.411 0.1581
## AlcQF.c -0.002117 0.009785 -0.216 0.8287
## BvN:AlcQF.c 0.004888 0.016353 0.299 0.7650
## BNvW:AlcQF.c 0.014897 0.014573 1.022 0.3066
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN BNvW AlcQF. BN:AQF
## BvN 0.008
## BNvW -0.142 0.181
## AlcQF.c -0.066 0.003 0.001
## BvN:AlcQF.c 0.003 -0.120 0.002 0.016
## BNvW:AlcQF. 0.001 0.002 -0.116 -0.241 0.366## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + Risk * BvN + Risk * BNvW + (1 + BvN +
## BNvW | Subject) + (1 + Risk | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5934.8 6031.0 -2952.4 5904.8 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.7414 -0.9357 0.5778 0.8355 2.3930
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.277214 0.52651
## BvN 0.668628 0.81770 0.02
## BNvW 0.521802 0.72236 -0.36 0.59
## pic2 (Intercept) 0.029692 0.17231
## Risk 0.008081 0.08989 -0.74
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.13253 0.08788 1.508 0.1315
## BvN 0.57187 0.17655 3.239 0.0012 **
## BNvW 0.24949 0.15416 1.618 0.1056
## Risk 0.14559 0.07702 1.890 0.0587 .
## BvN:Risk 0.03543 0.14363 0.247 0.8052
## BNvW:Risk 0.07996 0.12520 0.639 0.5230
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN BNvW Risk BvN:Rs
## BvN 0.017
## BNvW -0.158 0.203
## Risk -0.139 0.005 0.002
## BvN:Risk 0.006 -0.220 0.005 0.033
## BNvW:Risk 0.003 0.005 -0.221 -0.214 0.314## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + RAPI * BvN + RAPI * BNvW + (1 + BvN +
## BNvW | Subject) + (1 + RAPI | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5939.9 6036.0 -2954.9 5909.9 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.5182 -0.9471 0.5709 0.8410 2.3844
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 2.790e-01 0.528167
## BvN 6.561e-01 0.810026 0.03
## BNvW 5.199e-01 0.721025 -0.34 0.59
## pic2 (Intercept) 2.975e-02 0.172492
## RAPI 1.316e-06 0.001147 -1.00
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.132111 0.088044 1.501 0.13348
## BvN 0.569612 0.176030 3.236 0.00121 **
## BNvW 0.247332 0.154091 1.605 0.10847
## RAPI 0.018812 0.010024 1.877 0.06055 .
## BvN:RAPI -0.003381 0.017549 -0.193 0.84722
## BNvW:RAPI 0.001878 0.015426 0.122 0.90310
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN BNvW RAPI BN:RAP
## BvN 0.022
## BNvW -0.148 0.202
## RAPI -0.022 0.003 0.003
## BvN:RAPI 0.003 -0.033 0.002 0.036
## BNvW:RAPI 0.004 0.002 -0.038 -0.243 0.386## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ BvN + BNvW + AE * BvN + AE * BNvW + (1 + BvN + BNvW |
## Subject) + (1 + AE | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5941.1 6037.2 -2955.5 5911.1 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4612 -0.9414 0.5719 0.8440 2.3828
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 2.945e-01 0.542647
## BvN 6.243e-01 0.790129 0.00
## BNvW 5.169e-01 0.718949 -0.34 0.59
## pic2 (Intercept) 2.974e-02 0.172444
## AE 6.342e-06 0.002518 -1.00
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.131683 0.089361 1.474 0.14059
## BvN 0.568792 0.174632 3.257 0.00113 **
## BNvW 0.246758 0.153928 1.603 0.10892
## AE 0.006773 0.008962 0.756 0.44979
## BvN:AE 0.021602 0.015204 1.421 0.15538
## BNvW:AE 0.008169 0.013496 0.605 0.54496
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN BNvW AE BvN:AE
## BvN 0.007
## BNvW -0.147 0.198
## AE -0.042 0.000 -0.002
## BvN:AE 0.000 -0.079 0.001 0.016
## BNvW:AE -0.002 0.001 -0.073 -0.234 0.381This model includes all the questionnaire measures described above as fixed effects. Only intercepts were calculated for random factors due to convergence limitations.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula:
## response ~ BvN + BNvW + AlcQF + Risk + RAPI + AE + (1 | Subject) +
## (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5734.1 5791.4 -2858.0 5716.1 4278
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.0243 -0.9639 0.6044 0.8820 1.6984
##
## Random effects:
## Groups Name Variance Std.Dev.
## Subject (Intercept) 0.24057 0.4905
## pic2 (Intercept) 0.02882 0.1698
## Number of obs: 4287, groups: Subject, 63; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.113536 0.085183 1.333 0.182581
## BvN 0.500636 0.143300 3.494 0.000476 ***
## BNvW 0.196392 0.123775 1.587 0.112584
## AlcQF -0.015656 0.009796 -1.598 0.110010
## Risk 0.141897 0.086951 1.632 0.102696
## RAPI 0.018680 0.011280 1.656 0.097721 .
## AE -0.003393 0.009687 -0.350 0.726130
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) BvN BNvW AlcQF Risk RAPI
## BvN 0.004
## BNvW 0.002 0.005
## AlcQF 0.014 -0.003 -0.003
## Risk -0.018 0.004 0.002 -0.187
## RAPI -0.019 0.003 0.002 -0.340 -0.287
## AE -0.005 -0.001 -0.001 0.036 -0.432 -0.058Water vs Neutral
This model tests the effects of a Water vs a Neutral prime on risk response.
Fixed effects:
      Water/Neutral contrast: Water = +1/2, Neutral = -1/2, Beer = 0
      Orthogonal contrast: Water = +1/3, Neutral = +1/3, Beer = -2/3
Random effects:
      Subject: Intercept, slope for each contrast
      Stimuli(pic2): Intercept
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + (1 + BvW + BWvN | Subject) + (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5934.0 5998.1 -2957.0 5914.0 4482
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4593 -0.9455 0.5729 0.8396 2.3742
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.29754 0.5455
## BvW 1.02899 1.0144 -0.22
## BWvN 0.23781 0.4877 0.27 0.32
## pic2 (Intercept) 0.02976 0.1725
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.13198 0.08963 1.472 0.140900
## WvN 0.03759 0.16128 0.233 0.815692
## WNvB -0.55051 0.16567 -3.323 0.000891 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN
## WvN 0.147
## WNvB 0.053 0.112These models add questionnaire measures and interactions to the list of fixed effects and calculates a questionnaire measure slope for each stimuli picture.
Measures:
AlcQF - Alcohol use quantity-frequency product (drinks/month)
Risk - Risk taking after alcohol consumption
RAPI - Rutgers alcohol problem index
AE - Alcohol expectancy questionnaire (CEOA)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + AlcQF * WvN + AlcQF * WNvB + (1 + WvN +
## WNvB | Subject) + (1 + AlcQF | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5665.1 5760.6 -2817.6 5635.1 4272
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4848 -0.9404 0.5688 0.8360 2.4141
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 3.180e-01 0.56394
## WvN 3.598e-01 0.59986 0.41
## WNvB 7.606e-01 0.87212 0.14 0.37
## pic2 (Intercept) 3.702e-02 0.19242
## AlcQF 1.452e-05 0.00381 -1.00
## Number of obs: 4287, groups: Subject, 63; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.122011 0.096100 1.270 0.20421
## WvN 0.032401 0.174975 0.185 0.85309
## WNvB -0.515811 0.176049 -2.930 0.00339 **
## AlcQF -0.002118 0.009785 -0.216 0.82867
## WvN:AlcQF -0.012454 0.013858 -0.899 0.36880
## WNvB:AlcQF -0.011115 0.016398 -0.678 0.49787
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN WNvB AlcQF WN:AQF
## WvN 0.138
## WNvB 0.060 0.105
## AlcQF -0.066 0.001 -0.002
## WvN:AlcQF 0.001 -0.156 0.001 0.263
## WNvB:AlcQF -0.003 0.001 -0.094 0.095 0.218## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + Risk * WvN + Risk * WNvB + (1 + WvN +
## WNvB | Subject) + (1 + Risk | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5934.8 6031.0 -2952.4 5904.8 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.7414 -0.9357 0.5778 0.8355 2.3929
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 0.277219 0.52652
## WvN 0.338815 0.58208 0.47
## WNvB 0.769179 0.87703 0.14 0.36
## pic2 (Intercept) 0.029691 0.17231
## Risk 0.008081 0.08989 -0.74
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.13254 0.08788 1.508 0.131516
## WvN 0.03644 0.16134 0.226 0.821300
## WNvB -0.55366 0.16618 -3.332 0.000863 ***
## Risk 0.14559 0.07702 1.890 0.058725 .
## WvN:Risk -0.06225 0.12323 -0.505 0.613483
## WNvB:Risk -0.06655 0.14055 -0.473 0.635875
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN WNvB Risk WvN:Rs
## WvN 0.160
## WNvB 0.060 0.110
## Risk -0.139 0.001 -0.005
## WvN:Risk 0.001 -0.290 0.001 0.236
## WNvB:Risk -0.006 0.001 -0.177 0.070 0.169## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + RAPI * WvN + RAPI * WNvB + (1 + WvN +
## WNvB | Subject) + (1 + RAPI | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5939.9 6036.0 -2954.9 5909.9 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.5182 -0.9471 0.5709 0.8410 2.3843
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 2.790e-01 0.528167
## WvN 3.365e-01 0.580079 0.45
## WNvB 7.595e-01 0.871486 0.12 0.37
## pic2 (Intercept) 2.975e-02 0.172486
## RAPI 1.313e-06 0.001146 -1.00
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.132120 0.088042 1.501 0.133447
## WvN 0.037514 0.161280 0.233 0.816070
## WNvB -0.550865 0.165750 -3.323 0.000889 ***
## RAPI 0.018810 0.010024 1.877 0.060579 .
## WvN:RAPI -0.003569 0.014507 -0.246 0.805665
## WNvB:RAPI 0.001597 0.017637 0.091 0.927842
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN WNvB RAPI WN:RAP
## WvN 0.154
## WNvB 0.051 0.112
## RAPI -0.022 -0.002 -0.004
## WvN:RAPI -0.002 -0.052 -0.001 0.280
## WNvB:RAPI -0.004 -0.001 -0.026 0.079 0.218## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: response ~ WvN + WNvB + AE * WvN + AE * WNvB + (1 + WvN + WNvB |
## Subject) + (1 + AE | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5941.1 6037.2 -2955.5 5911.1 4477
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4611 -0.9414 0.5719 0.8440 2.3828
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Subject (Intercept) 2.945e-01 0.542642
## WvN 3.352e-01 0.578936 0.42
## WNvB 7.338e-01 0.856598 0.14 0.39
## pic2 (Intercept) 2.974e-02 0.172444
## AE 6.343e-06 0.002519 -1.00
## Number of obs: 4492, groups: Subject, 66; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.131689 0.089361 1.474 0.140569
## WvN 0.037660 0.161201 0.234 0.815278
## WNvB -0.549958 0.164551 -3.342 0.000831 ***
## AE 0.006773 0.008962 0.756 0.449791
## WvN:AE 0.002631 0.012720 0.207 0.836108
## WNvB:AE -0.020286 0.015303 -1.326 0.184964
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN WNvB AE WvN:AE
## WvN 0.145
## WNvB 0.063 0.116
## AE -0.042 0.002 0.001
## WvN:AE 0.002 -0.100 0.002 0.258
## WNvB:AE 0.001 0.002 -0.061 0.091 0.223This model includes all the questionnaire measures described above as fixed effects. Only intercepts were calculated for random factors due to convergence limitations.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula:
## response ~ WvN + WNvB + AlcQF + Risk + RAPI + AE + (1 | Subject) +
## (1 | pic2)
## Data: gamble
##
## AIC BIC logLik deviance df.resid
## 5734.1 5791.4 -2858.0 5716.1 4278
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.0243 -0.9639 0.6045 0.8820 1.6984
##
## Random effects:
## Groups Name Variance Std.Dev.
## Subject (Intercept) 0.24057 0.4905
## pic2 (Intercept) 0.02882 0.1698
## Number of obs: 4287, groups: Subject, 63; pic2, 12
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.113547 0.085181 1.333 0.182529
## WvN 0.053927 0.142689 0.378 0.705477
## WNvB -0.473671 0.124300 -3.811 0.000139 ***
## AlcQF -0.015655 0.009796 -1.598 0.110040
## Risk 0.141896 0.086951 1.632 0.102700
## RAPI 0.018680 0.011280 1.656 0.097720 .
## AE -0.003394 0.009687 -0.350 0.726043
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) WvN WNvB AlcQF Risk RAPI
## WvN 0.000
## WNvB -0.005 0.000
## AlcQF 0.014 0.001 0.003
## Risk -0.018 0.000 -0.004 -0.187
## RAPI -0.019 -0.001 -0.004 -0.340 -0.287
## AE -0.005 0.000 0.002 0.036 -0.432 -0.058Correlations
NIAAA Use Measures
These include raw and recoded (.re) values. P-values shown. 
Correlation coefficients shown. 
Combined
beer.dif = hi-risk beer responses - low-risk beer responses
water.dif = hi-risk water responses - low-risk water responses
neutral.dif = hi-risk neutral responses - low-risk neutral responses
AlcQF - Alcohol use quantity-frequency product (drinks/month)
Risk - Risk taking after alcohol consumption (alpha = .86)
RAPI - Rutgers alcohol problem index (alpha = .83)
AE - Alcohol expectancy questionnaire (CEOA, alpha = .74)
P-values shown. 
Correlation coefficients shown. 